Summer 1997

Strategies for telling truths

Visual Explanations: Images and Quantities, Evidence and Narrative

Edward R. Tufte<br>Graphics Press

The sharpest books on design come from small independent publishers, such as Hyphen Press in England and Edward Tufte’s Graphics Press in the U.S. For his first excursion into information design – The Visual Display of Quantitative Information, 1983 – Tufte set up this single-uthor publishing company. Within ten years the book and its successor (Envisioning information, 1990) had sold 170,000 copies. Each met with an almost universally uncritical reception and may be among the most lavishly praised books on design ever. Tufte’s new title, Visual Explanations, will be no exception.

This is partly due to the quality of all three books – the result of painstaking care and intellectual integrity. Tufte writes: ‘The books are meant to be self-exemplifying: the objects themselves embody the ideas written about.’ Necessity became a virtue, giving Tufte complete control over design and production. Elsewhere he has said of self-publication: ‘If I wanted to mess it up, I could go to a real publisher … It ill serves the wonders of information design to reproduce them poorly.’

To professionals whose work involves presenting information (architects, clinicians, meteorologists, engineers), Tufte’s books must come as a shock and a revelation. And to graphic designers used to their trade’s flim-flam, they may appear as if from another world, which they are. Edward Tufte is Professor of Political Science and Statistics (and Senior Critic in Graphic Design) at Yale University. He offers contemporary graphic designers few mentions and no peace, mocking their ‘obscurantist foolings around’ as a ‘triumph of decoration over information’, and dismisses, as ‘chartjunk’, the ‘decorative clichés of “info-graphics” … complete with celebrity factoids, over-compressed data, and Isotype styling’.

Tufte says that his first two books were about pictures of numbers and pictures of nouns respectively, while Visual Explanations deals with pictures of verbs. This metaphor is elaborated: ‘Assessments of change, dynamics, and cause and effect are at the heart of thinking and explanation. To understand is to know what causes provokes what effect, by what means, at what rate. How then is such knowledge to be represented? ... This book describes design strategies – the proper arrangement in space and time of images, words, and numbers – for presenting information about motion, process, mechanism, cause and effect.’

Visual Explanations is a tour of a ‘cognitive paradise of explanation, a sparkling and exuberant world’. It shows and analyses about 200 graphic examples, traces drawn from many domains of human activity: scientific visualisation, computer modelling, landscape design, oceanography, cosmology, optics, geometry. Tufte offers a close and critical reading of these examples, often contrasting ‘before’ and ‘after’ versions. He criticises the ‘dequantification’ of graphic aids to reasoning; thus ‘Arbitrary, transient, one-sided, fractured, undocumented materials have become the great predicament of image making and processing. How are we to assess the integrity of visual evidence?’

He laboriously documents conjurer’s tricks to make a point: ‘To create illusions is to engage in disinformation design ... the strategies of magic suggest what not to do if our goal is truth-telling.’ There are some unanswered questions here: a chapter on ‘Multiples in space and time’ shows the transformation of a patient’s medical records, from the typical data dump into a concise graphic overview: ‘Combining overview with detail, this one-page, high-resolution set of multiples makes sense of thousands of items now scattered throughout standard medical records.’ Yes; but this printed “page” – “high-resolution”’ in Tufte’s terms – shows a token for a ‘next screen’ button. This makes no sense: does the picture represent a screen or a printed page? It matters: such ‘high-resolution’ is impossible on screen displays. A similar elision occurs later in Tufte’s made-up ‘interface for guiding museum-goers’, which is also a printed page rather than a screen shot.

The most compelling chapter is on ‘Visual and statistical thinking’. It examines ‘the statistical and graphical reasoning used in making two life-and-death decisions: how to stop a cholera epidemic in London during September 1854; and whether to launch the space shuttle Challenger on January 28, 1986. By creating statistical graphics that revealed the data, Dr John Snow was able to discover the cause of the epidemic … In contrast, by fooling around with displays that obscured the data, those who decided to launch the space shuttle got it … terribly wrong. For both cases, the consequences resulted directly from the quality of methods used in displaying and assessing quantitative evidence’.

Snow’s map is meticulously reproduced here, and is fluently interrogated. Tufte then shows charts faxed by engineers to NASA just before the Challenger launch, observing a ‘scandalous discrepancy between the intellectual tasks at hand and the images created to serve those tasks’. He designs a table and a scatter plot graph with damage indexes for all shuttle launches before Challenger: both clearly show the risks of a cold weather launch. One chart laid before the investigative commission contained the information necessary to diagnose the relationship between temperature and damage, but in a form, (‘ludicrous and corrupt’ in Tufte’s words) that was barely usable. All this points him to an irresistible conclusion: ‘There are right and wrong ways to show data; there are displays that reveal the truth and displays that do not.’ For Tufte this might be obvious: just use your eyes.

Tufte is a true believer, but I am agnostic, and he encourages me to demand high standards of evidence. What, for instance happened between the ‘before’ and ‘after’ versions? Are there alternatives? Rich evidence of the process of designing – trials and errors, false starts and diversions – is missing. Then there are questions of judgement. Tufte’s ‘afters’ may pack in more detail, be clearer and more attractive. Are they therefore ‘better’? ‘Better’ for whom, for what purposes, in what circumstances of use? Tufte has not shown much interest in such questions, or in consulting users, in observing them using graphic materials, or in testing those materials.

Tufte has said that ‘the operating moral premise of information design should be that our readers are alert and caring; they may be busy, eager to get on with it, but they are not stupid’. Who would disagree? But he is stronger on the integrity of ‘information’ than on the diversity of audiences and readers’ purposes. Missing from Tufte’s many pronouncements is ‘know your readers: they are not you’. He knows better than most of us that scientific reasoning and interpreting statistical evidence are often counterintuitive. So in a future book I would like to see him try designing something for broad rather than specialised audiences: nutritional information on food packs, say, or instructions on medicine labels. I would like to see him show how the same information might be designed for different audiences asking different questions. And I would like to see more of his process of designing. Edward Tufte is well qualified for such a challenging project.

First published in Eye no. 25 vol. 7, 1997

Eye is the world’s most beautiful and collectable graphic design journal, published quarterly for professional designers, students and anyone interested in critical, informed writing about graphic design and visual culture. It is available from all good design bookshops and online at the Eye shop, where you can buy subscriptions, back issues and single copies of the latest issue.